>>Return to Tell Me About Statistics!
“Hypothesis testing” can be misleading because it uses a sample. It can be divided into two types, based on the error content.
In this article, we will discuss the “first type of error.”
Type I error
Hypothesis testing is performed to objectively confirm the effectiveness of the new drug. The null hypothesis is that the new drug would be as effective as the existing drug and the alternative hypothesis is that the new drug would be more effective than the existing drug.
When n = 50 and p = 0.04, the alternative hypothesis can be accepted from a p-value < 0.05; therefore, it can be said that the new drug is more effective than the existing drug.
The problem is that even though there is no difference in the efficacy of the new and existing drug (the null hypothesis is correct), the truth is overlooked, and the drug is judged to be effective. Such mistakes are called “Type I errors.” They are calledαerrors, and constitute about 5% of such mistakes (figure).
The probability of a type-I error (risk ratio) is denoted byα. This is the same as the significance level used for the test. “1-α” is the probability of correctly judging “no effect” as “no effect.”
[Figure] Test of significance level (two-sided)
>>Return to Tell Me About Statistics!
Comments are closed